English
Related papers

Related papers: Backtracking Improves Generation Safety

200 papers

Supervised fine-tuning has become the predominant method for adapting large pretrained models to downstream tasks. However, recent studies have revealed that these models are vulnerable to backdoor attacks, where even a small number of…

Cryptography and Security · Computer Science 2025-01-08 Peihai Jiang , Xixiang Lyu , Yige Li , Jing Ma

Efforts to ensure the safety of large language models (LLMs) include safety fine-tuning, evaluation, and red teaming. However, despite the widespread use of the Retrieval-Augmented Generation (RAG) framework, AI safety work focuses on…

Computation and Language · Computer Science 2025-04-28 Bang An , Shiyue Zhang , Mark Dredze

While there has been progress towards aligning Large Language Models (LLMs) with human values and ensuring safe behaviour at inference time, safety guards can easily be removed when fine tuned on unsafe and harmful datasets. While this…

Generative large language models (LLMs) have achieved state-of-the-art results on a wide range of tasks, yet they remain susceptible to backdoor attacks: carefully crafted triggers in the input can manipulate the model to produce…

Artificial Intelligence · Computer Science 2025-05-20 Yige Li , Hanxun Huang , Yunhan Zhao , Xingjun Ma , Jun Sun

We present InvThink, a training and prompting framework that requires the model to enumerate, analyze, and constrain potential failures before generating its final response. Unlike existing safety alignment methods that optimize only for…

Artificial Intelligence · Computer Science 2026-05-11 Yubin Kim , Taehan Kim , Eugene Park , Chunjong Park , Cynthia Breazeal , Daniel McDuff , Hae Won Park

Large language models (LLMs) have raised concerns about potential security threats despite performing significantly in Natural Language Processing (NLP). Backdoor attacks initially verified that LLM is doing substantial harm at all stages,…

Cryptography and Security · Computer Science 2024-07-09 Pengzhou Cheng , Yidong Ding , Tianjie Ju , Zongru Wu , Wei Du , Ping Yi , Zhuosheng Zhang , Gongshen Liu

Large Language Models (LLMs) have emerged as powerful tools, but their inherent safety risks - ranging from harmful content generation to broader societal harms - pose significant challenges. These risks can be amplified by the recent…

Currently, large models are prone to generating harmful content when faced with complex attack instructions, significantly reducing their defensive capabilities. To address this issue, this paper proposes a method based on constructing data…

Cryptography and Security · Computer Science 2025-01-03 Keke Zhai

AI developers often apply safety alignment procedures to prevent the misuse of their AI systems. For example, before Meta released Llama 2-Chat - a collection of instruction fine-tuned large language models - they invested heavily in safety…

Machine Learning · Computer Science 2024-05-24 Simon Lermen , Charlie Rogers-Smith , Jeffrey Ladish

Large language models (LLMs) have achieved impressive performance in code generation recently, offering programmers revolutionary assistance in software development. However, due to the auto-regressive nature of LLMs, they are susceptible…

Software Engineering · Computer Science 2025-03-25 Xue Jiang , Yihong Dong , Yongding Tao , Huanyu Liu , Zhi Jin , Wenpin Jiao , Ge Li

Backdoor attacks compromise model reliability by using triggers to manipulate outputs. Trigger inversion can accurately locate these triggers via a generator and is therefore critical for backdoor defense. However, the discrete nature of…

Cryptography and Security · Computer Science 2026-03-18 Ruyi Zhang , Heng Gao , Songlei Jian , Yusong Tan , Haifang Zhou

Large Language Models (LLMs) can generate code but often introduce security vulnerabilities, logical inconsistencies, and compilation errors. Prior work demonstrates that LLMs benefit substantially from structured feedback, static analysis,…

Cryptography and Security · Computer Science 2026-01-05 Vidyut Sriram , Sawan Pandita , Achintya Lakshmanan , Aneesh Shamraj , Suman Saha

While large neural-based conversational models have become increasingly proficient dialogue agents, recent work has highlighted safety issues with these systems. For example, these systems can be goaded into generating toxic content, which…

Computation and Language · Computer Science 2023-10-24 Nicholas Meade , Spandana Gella , Devamanyu Hazarika , Prakhar Gupta , Di Jin , Siva Reddy , Yang Liu , Dilek Hakkani-Tür

Aligned language models face a significant limitation as their fine-tuning often results in compromised safety. To tackle this, we propose a simple method RESTA that performs LLM safety realignment. RESTA stands for REstoring Safety through…

Computation and Language · Computer Science 2024-02-20 Rishabh Bhardwaj , Do Duc Anh , Soujanya Poria

Large Language Models (LLMs) are typically trained to predict in the forward direction of time. However, recent works have shown that prompting these models to look back and critique their own generations can produce useful feedback.…

Computation and Language · Computer Science 2025-02-04 Yerram Varun , Rahul Madhavan , Sravanti Addepalli , Arun Suggala , Karthikeyan Shanmugam , Prateek Jain

Because state-of-the-art language models are expensive to train, most practitioners must make use of one of the few publicly available language models or language model APIs. This consolidation of trust increases the potency of backdoor…

Cryptography and Security · Computer Science 2023-07-28 Nikhil Kandpal , Matthew Jagielski , Florian Tramèr , Nicholas Carlini

Fine-tuning large language models (LLMs) on additional datasets is often necessary to optimize them for specific downstream tasks. However, existing safety alignment measures, which restrict harmful behavior during inference, are…

Computation and Language · Computer Science 2024-10-15 Minjun Zhu , Linyi Yang , Yifan Wei , Ningyu Zhang , Yue Zhang

Large language models (LLMs) are typically aligned to refuse harmful instructions through safety fine-tuning. A recent attack, termed abliteration, identifies and suppresses the single latent direction most responsible for refusal behavior,…

Computation and Language · Computer Science 2025-10-08 Harethah Abu Shairah , Hasan Abed Al Kader Hammoud , Bernard Ghanem , George Turkiyyah

Large language models (LLMs) have acquired the ability to handle longer context lengths and understand nuances in text, expanding their dialogue capabilities beyond a single utterance. A popular user-facing application of LLMs is the…

Computation and Language · Computer Science 2024-10-29 Terry Tong , Jiashu Xu , Qin Liu , Muhao Chen

Large language models deployed as agents increasingly interact with external systems through tool calls--actions with real-world consequences that text outputs alone do not carry. Safety evaluations, however, overwhelmingly measure…

Artificial Intelligence · Computer Science 2026-02-20 Arnold Cartagena , Ariane Teixeira